Like the way you make HA visible in the format of lovely cats.
An open-source MySQL alternative for running database on Kubernetes
41–50 of 61 posts
Re: An open-source MySQL alternative for running database on Kubernetes
#42I like the message it conveyed 'No need to be a K8s professional'. That is exactly how I feel after creating a mysql cluster smoothly. I tried oracle mysql opertor months earlier, but failed to install it on AWS EKS without knowing why.
Re: An open-source MySQL alternative for running database on Kubernetes
#43If this is supposed to be used in production as open source, then I am a little surprised that it's under AGPL 3.0, especially given there's no mention of commercial version.
Depending on one's needs and interests, they changed it from Apache 2 a few months ago: https://github.com/apecloud/kubeblocks/blob/5ba9b39716940cd7... I didn't dig into the changes to see if there's anything super interesting that happened between then and now
It really, really jams me up that the GitHub license thing is so stupid that when looking at a file containing the Apache 2 license it continues to say AGPL right across the top of that file. It's so misleading. Their license sniffer always shows $default_branch/LICENSE regardless of what branch (or file) one is looking at
Re: An open-source MySQL alternative for running database on Kubernetes
#44Earlier quoted context omitted.
So... throw money on compute instead of just adding an index..?
I’m offended that you don’t take even 10 minutes to understand why this will not work. People say they don't need sysadmins and stuff like this is the result. Vertical scalability limits are very soon met and can be difficult to fix post-fact.
My scepticism was clearly not well enough signalled, apologies.
To be clear I agree "waste money to avoid applying basic database techniques" is A Bad Thing!
Re: An open-source MySQL alternative for running database on Kubernetes
#45Earlier quoted context omitted.
That is the reason why you need KubeBlocks, it takes over the complexity.
That is not how complexity works. You might lower the difficulty in the happy path, but when something goes wrong, there is additional complexity, and that means more places to debug and understand.
This is my primary objection to 3rd party kubernetes operators. I really don't care how hard something is to setup or learn to manage. I care how hard it is to recover, tune, and scale.
Re: An open-source MySQL alternative for running database on Kubernetes
#46This "why you need KubeBlocks" reads like a bunch of pointless marketing speak (quoted below).
" >Why you need KubeBlocks When adopting a multi-cloud or hybrid cloud strategy, it is essential to prioritize application portability and use software or services that offer consistent functionality across different infrastructures."
I'd rather read some concrete examples or use cases.
Re: An open-source MySQL alternative for running database on Kubernetes
#47As we know, although Kubernetes (K8s) can be used to build and deploy various applications, including databases, there are still some challenges and limitations related to databases that cannot be fully addressed. Here are some of these issues:
1. Data Consistency: Kubernetes does not directly provide guarantees for data consistency in databases, while databases typically require integrity and consistency of data. This may rely on mechanisms such as database replication, sharding, and transactions to ensure data consistency, rather than solely relying on Kubernetes.
2. Data Persistence and Performance: While Kubernetes provides solutions for persistent storage, such as Persistent Volumes (PVs), it may still not fully meet the requirements of high-performance databases, such as large-scale data processing or high-concurrency workloads in relational databases.
3. Distributed Deployment and Maintenance of Databases: Distributed databases often require data synchronization and load balancing among different nodes. Kubernetes can provide partial support, such as StatefulSets for managing stateful applications, but the distributed nature and complexity of databases may require additional configuration and management to ensure data consistency and high availability.
4. Database Scaling and Fault Recovery: In large-scale applications, databases may need horizontal scaling and fault recovery capabilities. While Kubernetes can dynamically scale and manage containers, scaling and fault recovery at the database level may require more complex configuration and strategies, depending on the specific database engine and architecture.
5. Database Backup and Recovery: Backup and recovery are critical tasks in database management. Although Kubernetes provides snapshot and backup functionality, the backup and recovery process for databases may require additional considerations, such as handling transaction logs and ensuring data consistency.
While Kubernetes provides some convenience and abstraction layers for building databases, the above issues indicate that specialized database architectures and solutions are still required for specific database requirements and scenarios to meet the needs of data consistency, persistence, performance, and scalability.
I have been downloaded kubeblocks and running it on local k8s, AWS EKS, GCP GKE, it can create many kind of databases very quickly. Also, kubeblocks includes day2 operations keep me free from tons of DBA jobs.
I will keep trying kubeblocks. When the day it meet the needs of data consistency, persistence, performance, and scalability, go production and take a good sleep.
Re: An open-source MySQL alternative for running database on Kubernetes
#48Earlier quoted context omitted.
And that guy is usually me. Just remember when you leave a pool of shit behind, one of the victims might be making a hiring decision about you at some point in the future.
Read that last part again folks. Twice if you have to. You will run into the same folks in the industry if you last longer than 10 years.
“The industry” extends far beyond and is more than the big tech circle jerk. I’ve bounced around outside of the big tech circus for nearly 15 years now and have -never- unwittingly run into someone I previously worked with. Have I targeted institutions that I know employ people that I've crossed paths with in the past? Yes, deliberately for one reason or another, but never once have I been hired, paged through a company directory and been surprised to the point of "oh, they're here?".
Re: An open-source MySQL alternative for running database on Kubernetes
#49Earlier quoted context omitted.
So... throw money on compute instead of just adding an index..?
I’m offended that you don’t take even 10 minutes to understand why this will not work. People say they don't need sysadmins and stuff like this is the result. Vertical scalability limits are very soon met and can be difficult to fix post-fact.
It's phrased as a joke, no?
Re: An open-source MySQL alternative for running database on Kubernetes
#50After looking at the readme, I'm still unclear on what kind of product it is - another database operator? In the readme I don't see any advantages it has compared to other operators, why not I choose other more mature operators?
Are the "Goals" serious? That looks more like empty slogans. Perhaps it would be better to explain in detail what it have actually done and how did it.
Furthermore, when can change that misleading title?